Golo Föllmer: Soft Music


This essay combines three layers of ideas:

The first one is in text form, containing personal reflections on the possible creation of a new kind of music by means of a network. These means are technical as well as communicative, instrumental, conceptual and perhaps even musical.

The second layer is a collage of thoughts collected from musicians and artists working on the Net. Presented as video clips, they reveal the humans behind the ideas. These clips are juxtaposed with or illustrate my own reflections.

The third layer consists of hypertext links to sound projects, and, if such projects are not accessible, to documentation and the like.

The general idea of the essay is not solely to be found in the text layer, nor in the choice of links or in the artists’ statements. While it is certainly fine to stroll through each of the separate layers, the subject’s diversity is more closely captured when combining multiple approaches and views. The overall idea of the essay can be discovered in the combination of the three layers. The fact that there is no one single path through the layers, but instead, a multiplicity of paths, reflects perfectly on the nature of the topic: the network as a multi-dimensional space of unstable musics; of musics in flux.

Many thanks for deeply inspiring thoughts to Josephine Bosma as well as to Guy van Belle and the other artists who physically and mentally supported the conversations documented here — also, to Peter Traub who couldn't be included in the videos for technical reasons. More gratitude for sincerely valuable support to Christiane Hausmann, wonderful Claudia Mattern, Tiri Ciri International and the ever great Wolf and Mary Föllmer. And a broad thanks to all the participants of the various lists and live discussions on the subject, and finally to all those, who act musically on the net.


Is it musical?

The talk of the moment is that MP3 and Napster are causing upheaval in the music industry. Meanwhile, broadcasting via RealAudio opens the possibility of the computer as both receiver and transmitter. Music is heard over both formats for different purposes, according to the quality and speed provided by the format. Thus, audio formats change the possibilities and practice of music distribution as well as the way music is used.

These are important questions, however, I want to discuss a different one, one that is usually left out as if it had no importance: what does all this fuss about transmission, sound quality and public accessibility have to do with musicality; with a musical use of a technology? What makes a medium musical?

Thinking of older audio media — radio, telephone and recording technology, in general — it is hard to point out exactly which aspects make them more or less musical. Generally speaking, a combination of technical possibilities and extended human experience in the use of a medium are the criteria. In the end, what accounts for actual musical developments is, of course, incalculable. Is it more a way of instrumental use, or is it aspects of sound quality and diversity? Is it the extensive dialogue with a musical tradition or the intensive dialogue between musicians?

Is the network a musical medium?


So, I will not discuss aspects of distribution, because I believe that there is no new musical experience in listening to Stravinsky, Aphex Twin or my neighbour’s band over the Net. Instead, I want to discuss new approaches to making music that carry the qualities of a network: musical patterns of inter-connected digital environments that we have not heard before.

So let’s skip-click through some ingredients of what could make for musicality. Any instruments out there?


What is an instrument?

A musical instrument produces sound waves, reaching human ears mostly as vibrations of air. Currently, we are still locked in this principle, since no one yet knows how to directly infiltrate the auditory nervous system. However, most other ingredients of what constitutes a musical instrument have changed drastically in the networked computer.

First of all, the advantage of the computer, in comparison to traditional instruments, lies in its ability to create an infinite sonic universe by means of a multitude of sound synthesis techniques: imitations and extensions of physical instruments, digital emulations of analogue synthesis methods and inventions of new principles only realizable in the digital domain.

On top of that, digital sampling yields more possibilities than the traditional analog recording “sampling” technologies of LP’s and magnetic tape. The computer, however, is not as good in imitating traditional instruments. A main reason for this remains the clumsiness of computer interfaces for sound modulation, a technique accomplished by refined gestures in traditional, physical instruments.

»computer music«

So these are the basics that the computer alone provides. With the introduction of networks, however, this infinity of actual and possible instruments can now be located, gathered, run, understood and finally, utilized by every network user at the speed of the his/her brain. Every existing software instrument is readily accessible to the user on the network, with the exception of those which are copy protected through commericial schemes. It is a milieu of microscopic niches, where new styles and concepts flourish by themselves with little institutional guidance. The change occuring here introduces a new do-it-yourself culture.





Flexible Instruments

Taking a closer look at the functionality of musical instruments reveals several changes. One change is that music develops into more flexible structures by means of special networked instruments. While a musical composition used to be a fixed series of sucessive auditory elements in time, it now becomes a field of elements freely navigable and thus, with an infinity of possible realizations in time. The process of composition moves away from the notation of timed events in a score towards the design of an instrument. Audio capabilities, design of the user interface and inlets/outlets from and to the network set up a frame of possibilities. The instrumental structure sufficiently describes such music.

This general concept has found a multitude of realizations on the Net. Some of these focus on a conceptual idea, making specific structures or content of the medium audible.


Many of these actually don’t use the network to make the music, but only as a distribution channel for a musical tool. These works could also be found on a CD-ROM. Their connections to the Net are sometimes only marginal, like the “contact” button or the possibility to save a personal realization of the piece, in fact a score — which in itself is a kind of conceptual setback.

However, even if these instruments could also be distributed without the Net, the fact is that most of them would not be. They are small games, suitable only for distribution at no cost. Some are real instruments that offer complex controls and refined sounds. Additionally, there are systems that follow the idea of hypertext to control the flow of music in a non-linear way. Others are more suitable for providing background ambiences for the living room. Still others are more like toys, focusing on little ideas and giving the player the chance to try out and practice a highly specific skill in manipulating flexible music.

»Lexikon Sonate«
»Bruce Lee«
»Tilman Kuentzel«
»Absolut DJ«



MUM -Multi User Music

A variety of different instruments can be found. One model, for example, allows for a piece to be successively composed by a number of different people. Sergi Jordà’s software FMOL provides a two part environment in which to do this: a stand alone instrument (program) run on a local computer (Windows 95/98 platform) and a server-side database where pieces and improvisations played on the host computer can be stored.

The crucial point is that storage is not done in an audio file format but rather in a format that describes how an individual user has configured and acted on the FMOL instrument over the time span that s/he has decided to record. If published on the database, this “piece” can then be listened to and transformed by other users. Restrictions insure, however, that some aspects of the originator’s decisions are kept, while others can be altered, replaced, or supplemented.

What evolves out of this is the idea of a generative evolution of music, with parents and different numbers of generational offspring. It is interesting to follow the lines of development, compare generations and find personal styles in different transformations of such music. It is an aural equivalent of an evolutionary principle.



»Radiant Musiek«


MUI -Multi User Instruments

There is yet another possibility opened up by Networking. If war games can be played with several persons logged into the same game at the same time, then such multi-user ability should be possible with a musical instrument as well.

A minimalistic but effective example of such Multi-User Instruments is the performance NetOsc by Sensorband. The instrument consists of three sine wave oscillators that can be detuned in fine steps. Each of the sine tones is controlled by a different performer in a different location, but everyone hears all three tones together. Graphical elements show control change in realtime. The audience watches these controls on a screen. The music results from a tension between audible sound movements and visible control movements.



Another approach to this idea is Chris Brown’s as yet unfinished “Eternal network Music Site.” Brown envisions a Web site, where one or several flexible musics are permanently going on. Whoever passes by can join in to play on the accessible controls of the music system, interacting with the machine as well as with other players. Unlike the more conventional forerunner of this concept, the MIDI environment “Rocket Network,” Brown envisions a new way of listening to music evolving here — a hybrid between listening and playing.

»Rocket Network«
»eternal music«


CUI - Collective User Instruments

Pushing the distributed character of musical instruments one step further leads us to the point where several people not only play together on one instrument, but design it collectively. This can be done with a number of sophisticated software applications, two of the most influential being Max/MSP and Supercollider. Both programs play an immense role in the development of musical ideas, because they provide the framework for new concepts of music involving flexible processes.

Furthermore, users of these environments discuss technical and aesthetical questions on mailing lists. Feedback from these discussions is collected and used for future program development. The result of this process is that such environments that enable collective instrument design are also designed to some degree collectively.



Max/MSP and Supercollider both support the OpenSoundControl protocol, enabling machine users to construct complex connections between dislocated musical systems. As musicians design their instruments they can easily allow incoming data from other musicians to have control over certain aspects of their “patch” or self-made instrument. Such patches can easily be exchanged, so that several musicians can share the same instrument running on their computers. Individual alterations to a patch’s design can then be easily distributed to the other members of an instrumental design group, and so on.




What is next?

A stumbling block in the design of interactive instruments is their clumsy control interfaces. This is especially true with the concept of musical toys where input devices are restricted to mouse and keyboard, since today the instrument designer cannot ensure that the home user has access to other control devices. There is hope that this may change with newer devices that have more intuitive user interfaces but past research in this direction hasn’t been succesful and there is scepticism whether something really new can be developed. Former interfaces like the Matell PowerGlove, for example, proved to attract limited attention.

However, some experiments demonstrate that even with a mouse and keyboard, there is a chance to control an instrument in a rather fluid, musical way. The aforementioned “FMOL” can be played even by beginners with surprising complexity, using mouse and keys simultaneously. At the same time, it can still be interesting as a playing tool for someone with more sophisticated electronic music skills.


The next important direction for networked instruments seems to be that of mobile code. Currently, many applications only run on specific platforms (many on Windows but far more on Macintosh). The development of Java established a cross platform language that can run inside any Web browser. Porting existing as well as new applications onto this mobile platform on a larger scale however is only a recent development. Jsyn, for example, is an environment deeply inspired by the aforementioned Supercollider, providing the basis for two lines of development. First, it gives a broader public accessibility to the networked concept of musical system design, allowing many more users to work on complex musical processes. Second, Jsyn is a chance to spread the concept of process-based music more widely, since with mobile code it can be included in every Web site, just as animations and limited audio currently can be done in Flash, Shockwave or Beatnik.

»Java Music«


Musicians and Listeners

Looking at the changes musical instruments are going through gives rise to the impression that there is a radical shift taking place. Will listeners replace the musical performer in some sense? Might listeners even take over considerable parts of what used to be the composer’s responsibility? In other words, are we witnessing the dissolution of the musician?

Without a doubt, the audience finds itself in a new situation. The very fact that music for an active audience needs to suit different situations and attention spans, results in flexible music.

Describing the role of the musician cannot ignore that s/he is pulling back. S/he is, in fact, handing the conductor’s baton to the user of an interactive music system.

Yet, on the other hand, the musician receives several things in exchange, the main one being the fact that s/he now controls the whole production process. Whereas in the past the composer’s role was mostly confined to depicting a musical sensation most accurately in order to have others choose an environment for the presentation and production of the sounds, s/he has now taken on these tasks as well. Moreover, s/he is also responsible for the new element of the interface to the listener or user.

Yet, one can argue that the concept of planned tension and relaxation has not yet lost all of its importance for music and, if so, that there should be a means by which to create, control or at least, to trigger such processes. Technical interfaces may be one way of providing this. They are meant to translate concepts of dramaturgic structure into a flexible form. The artists’ idea of form, of corresponding changes of sound in time, needs to be programmed into the interface for it is this interface that represents the music.


Technical Resources

The idea of interactive interfaces between a musical work and an audience is gripping. However, upon further study, interfaces prove to be better at posing new problems rather than delivering instant solutions.

The promise that such interfaces give is the fascinating option that each member of an audience can design his/her own realization of a musical work, adjusting the resulting music based on individual knowledge, interest and other factors that differ from person to person.

This includes the hope that a deeper connection to the musical work is established, since more freedom is assumed to lead to deeper involvement. This is interesting, since this involvement catapults the audience into a position where they can act on a work, possibly expressing corresponding thoughts beyond the scope of the originator. For a musician, the potential of interactive music is that the audience adds something to it; that the participation of the audience makes a piece of music develop, evolve and take on a life of its own.

Technology offers one way of realizing this goal, yet there are major obstacles to this approach. A major problem grows out of technical limitations: high quality audio cannot easily be streamed in realtime, while the MIDI sounds found in most users’ computers sound tacky. More effective control interfaces will be developed, but control is mainly bound currently to the clumsy mouse and keyboard. Another problem is the difference in musical skill levels among users. This frequently leads to modes of interaction that are merely reactive (as predictable as a light switch) rather than interactive (to some degree, unpredictable and thereby, more organic).

»working group«





Human Resources

We come closer to the concept of a music taking on a life of its own as the range of freedom for the user grows; collaboration, perhaps more appropriately than interaction. This is, of course, the case when people develop systems of musical communication collectively. It is, however, also the case when there are further possibilities for a user to try tools, materials or ideas offered in an online music project.

Yet such ideas and opportunities don’t actually have to be offered; they can simply be found or taken. Napster proves this and the activities for copyleft make it into a philosophy — the appropriation of technological and intellectual resources for the sake of their further development.

»kainn untitled«


People Change

After all, it appears certain that roles of those concerned with the making and the listening of music are changing and developing new contours.

One role is that of the musician who makes his/her own instrument in software and in this process, utilizes the Net to gather tools and knowledge, discuss aesthetic ideas and exchange aural realizations of new instruments — a do-it-yourself approach to musical instruments.

Another role is that of the composer who plays the instrument which s/he “composed” by him/herself — a composer-performer. This model is certainly not new, but rather inherited from experimentalists like David Tudor.

Listening is becoming a more casual activity. This is due, in part, to the attention paid in playing, controlling and understanding the whole system of visual, contextual and other elements in addition to purely sonic ones.

At the same time, as the listener becomes more active, do we get something like a listener-performer?

It could be said that the new listeners, like an audience in a traditional concert setting, are still only putting elements together; in the past by way of their brains and now, additionally, on their screens. Yet, as in the past, they do it only for themselves since they are not presenting, performing or mediating music!

So we may have to think about a more open way to define the notions of the composer, the performer, the listener. The composer would (still) be a person who sets a scope and lays connections of ideas, concepts and atmospheres in music. The performer would be the person to mediate a primarily auditory event, albeit following different rules than in the past. Finally, the listener would be the one to put it to a final order, to receive it at a final destination and interpret his/her personal version of it. Whatever will happen, the main effect is that the range of roles for each grows.


What is music?

First, whatever sound we desire to call musical or noise is to some degree our personal choice; the rest, a matter of cultural agreement or tradition. In Europe, for long periods of time the sounds of highly sophisticated instruments were practically the only sources of musical sounds, wheras other cultures have included all sorts of found sounds in their music. Today, the western notion of musical sound has indeed broadened. Machine noise or a bee’s hum can easily be heard as musical if we focus our mind on the idea of music.

Second, the question of what structure of sounds we desire to call music deserves the same answer: it is a matter of personal and traditional values. So music is not something existing independent from us. Just as a (good) musician does not reproduce an independently existing music but makes a music of his/her own mind happen, a (good) listener does not simply receive an existing music but makes a music of his/her own happen in the mind during the individual process of perception. Music is a state of mind.

As minds are different and as minds change there are differing ideas around concerning which mental attitudes can be successfully involved in music. A major divide to be found here is the one between European art music, on one side, and popular as well as much of Non-European music on the other. Generally speaking, popular and, in fact, most ethnic musics focus on musical practice as a collective social activity (a game or rite) or means of expressive narration. Contrary to that the European art music tradition has to be thought of as a permanent juxtaposition of emotional and rational elements, and the corresponding state of mind is thus always connecting emotive sense and analytical reason.

What about the network state of mind?


Repetition and no repetition

Repetition is a central characteristic of the network and it is a central technique of computer- and network-based music: the role of looped sounds as a blend of instruments, using melodic and rhythmic elements. While the sounds employed may be complex and possibly hard to distinguish (as opposed to the well known instruments of the orchestra or the pop band), they establish contingency by being repeated, layered and only gradually modified. This is especially true in the network, since narrow bandwidth forces people to reduce the size of audio files. Reusing a file over and over is therefore a very data-efficient use, and most projects in some way loop their sound continuously. Repetition rules.

On the other side, repetition is totally out! What most people usually do with music they like is listen to it over and over, to the point of sensory saturation. The speciality of networked music, however, appears to be the opposite: flexible processes avoid identical repetition. Networked music embodies the idea of an everyday music and musical practice that is closer to daily life experience rather than the older idea of a refined art work, fixed by means of notation or recording.




As there is no physicality to be experienced in the network, three of four dimensions are absent. We are left solely with time. Furthermore, as there are few opportunities to synchronize to the time of others, as we do in a room with others or on the telephone, there is mostly our own time.

The change in music invoked by the specifics of the Net, basically a shift from a preconfigured flow of elements in time to a process-based system, means that we don’t know where we are in music. We are at one place at one time and that time is now. This increase of now-ness, as Josephine Bosma calls it, is not unique to networked music. It began with the concept of sound installation, where the momentary perception of sounds in the listener often had more importance than the overall order of elements in time or space.

The change in music invoked by the specifics of the Net, basically a shift from a preconfigured flow of elements in time to a process-based system, means that we don’t know where we are in music. We are where only we are and the time is: now. This increase of now-ness, as Josephine Bosma calls it, is not unique to networked music. It began with the concept of sound installation, where the momentary perception of sounds in the listener often had more importance than the overall order of elements in time or space.

Surprisingly, while there is only the “now,” small deviations from this narrow time frame prove most interesting for musicians to experiment with. Networked installations like Summer and many projects using RealAudio streams reflect the tension of time deviations — the quality of this unfocused now — into an audible structure. In fact, asked about sound on the Net, many musicians mention its unpredictable time delays. Daily experiences, like tiny temporal glitches and waiting periods, similar to successive arrivals of page content, yield small holes in an audio stream and other often beautiful artifacts which are turned into a musical experience.

»Sound Drifting«


Social space vs. physical space

If there is no physical 3-D on the Net, is there a space at all? Some projects play with the connection of virtual and physical space. For Global String, Atau Tanaka set up two strings in different locations to be plucked by gallery visitors. By playing the strings, sound is transfered and heard at the other site, which later comes back like an echo. Similar approaches are followed in a range of projects.

»Global String«

Yet, if I confine myself to the pure data structure of the network, there is 2-D on the screen for navigation, and there are n dimensions for the infinite amount of possible representations of linked ideas. These, however, are merely functional or abstract, i.e., intangible. What is tangible to the user is his/her own activity. It is navigation and communication in data space, as well as perceiving other users or processes navigation and communication which maps such a space.

»cloud chamber«

While there is no space in the usual sense, places are still established — not by someone putting something somewhere, but by active use; individually assigning meaning to things and ideas found in the data. This is basically how anthropology defines “place” in the physical world, but it is just as appropriate for the digital realm.

Accordingly, data space can be described as an anthropological or social realm — human activity is what establishes value and thereby, place, within it. If access into this data space remains open (i.e., if the necessary software tools are freely available and the channels for communication are provided), it can offer a democratic structure. There are a number of projects experimenting with the musical nature of these characteristics as different ideals of democratic processes are ported into music.

»Bits & Pieces«

If you want democratic principals to rule music, you get game-like events that metaphorically show aspects of social interaction. These might be intriguing to follow concerning the structures shaping them, concerning the powers acting in them and concerning their results.

Such socio-anthropological music proves to be interesting to take part in, but for an audience it is only compelling to watch or listen if the processes are, to a degree, transparent and understood. The change in the making of this music has to be accompanied by a change in listening. It doesn’t make sense to seat an audience in the old manner of being at the concert hall. Either everyone participates or measures have to be taken to translate what is going on in musical communication to the listeners.


Wild thinking, wild music

Years ago, working with computers used to be associated with rigid structures and strict rules. Yet experience with networks has made it appear as if this is not the best way by which humans can incorporate the medium into their lives — and vice versa. Research in artificial intelligence today follows the connectionist approach, assuming that intelligence — and musical creativity as a part of that — cannot be achieved by constructing large sets of rules, but can instead emerge from complex systems of autonomous agents. This philosophy is paralleled in the widespread notion of “soft” computer programming involving processes of trial-and-error.

“How shall I know what I mean before I hear what I say?” This paraphrase from Alice in Wonderland gets to the heart of the matter. Yet it also gets to the heart of music, since music is always linked to its physical realization. If I only imagine music, for instance, when reading a score, I imagine hearing it in my head.

Following this thread, what Claude Lévi-Strauss describes as “wild thinking” is how much musical experimentation on the Net is being done. Musicians and artists are not asking themselves which auditory results they are aiming at, which structure of material is necessary and consequently which tools are needed. This would be the good old western way of analytical thinking. Instead, they use material and tools found at place, modifying and recombining them — a process of thinking acted out with concrete materials at hand or an associative type of thinking described as bricolage.



This influence of connectionist ideas is not only present in artistic methods. It reaches over into musical results. The network state of mind shapes the idea of an emergent musical creation; the idea of a “soft” music.



What is the problem?

Speaking about networked music proves to be problematic. One problem is that most reflection on audio experimentation on the Net is done from the perspective of the media arts and far too little from a musicological viewpoint. The result of such reflection, while certainly yielding deep thoughts on changes in cultural practice, conceptual issues, media use and collaborative development of projects, is that aesthetic issues are mostly avoided. It is as if those in the media arts were not interested in reflection about music, but rather about experiencing it primarily as an event of social interaction with music in the background.

The other problem when speaking about networked music comes from the opposite side. Those with a contemporary music background who are trained to reflect on inner-musical relations usually react to audio experiments on the Net with comments like: doesn’t sound interesting, there are not enough structural / formal ideas to be heard, sounds resulting from the interface are boring, etc. It seems as if this group was not interested in reflecting about changes in cultural practice as crucial elements of networked music, but instead experiencing it solely as auditory events with a mouse attached.

Yet, networked music is an amalgamation of these two elements and in this it develops a weird mix: it can be playful and immediate as well as densely conceptual and mediated. On one hand, it can be explicitly open, promoting public access on and on the other, hermetic and highly developed technically, aesthetically and socially. The reason for this mix is that unlike most other cultural developments, it brings together an eclectic combination of people and ideas from popular and high-brow culture, from music and the fine arts.

Although musical experimenting on the Net will come into its tenth year soon, it is hardly a music of its own yet. This is certainly not surprising. It took over 50 years of experimentation with recording technology (the grammophone) to lead to acknowledged and influential musical results with Musique Concrète. Similarly, with networked music, we presently are talking about potentials of music which appears sometimes strange, sometimes awkward, because it requires a new musical state of mind in order to be made and to be heard.

What can we do? Use technology to reflect music. Use music to reflect technology. Listen to it. Talk about it. Try it out. Make your wishes.